Unmanned Aerial Vehicles Collision Avoidance from Moving Obstacles

The usage of the unmanned aerial vehicles (UAV) is becoming more diverse day by day in both military and civil applications. Specifically in military applications, they are becoming a "must-have" component in every arsenal. Low cost, and especially the importance of human life, makes these vehicles desired in any military situation. But in order to take advantage of the UAVs, these vehicles should be able to navigate safely and not collide with other air vehicles, especially ones that are manned and carry personnel. As a result, the importance of a collision avoidance system of the UA V arises and begins to play an important role from among other considerations and specialties. The See-and-A void method is the approved technique used for collision avoidance for any aerial vehicle. While using this method, the UAV gets data about the environment by using sensors and sends these data to the operator at a ground station. After downloading the data, the operator makes an evasive movement if necessary. But this method has drawbacks especially for the UAVs. For example, there can be data link problems while sending the required data to the operator. The emerging method for collision avoidance is Detect-and-A void. There are a lot of on-going studies about this method. With this concept, the UAV can be fully autonomous. It will sense the environment and can avoid any possible collision very quickly because UAVs react faster than human operators. In this thesis, a predictive algorithm for collision avoidance from moving obstacles is studied for the Detect-and-Avoid concept. The UAV is capable of predicting any potential collisions with moving obstacles and can make an evasive movement to avoid the collision. The algorithm is simulated by using MATLAB program. By using MATLAB, several scenarios are shown to describe how the algorithm works and tests its validity.